First, then, the life of the Palæozoic was remarkable, as compared with that of the present world, in presenting a great prevalence of animals and plants of synthetic types, as they are called by Agassiz that is, of creatures comprehending in one the properties of several groups which were to exist as distinct in the future. Such types are also sometimes called embryonic, because the young of animals and plants often show these comprehensive features. Such types were the old corals, presenting points of alliance with two distinct groups now widely separated; the old Trilobites, half king-crabs and half Isopods; the Amphibians of the coal, part fish, part newt, and part crocodile; the Sigillariæ, part club-mosses and part pines; the Orthoceratites, half nautili and half cuttle-fishes. I proposed, in the illustration in a former article, to give a restoration of one of the curious creatures last mentioned, the Orthoceratites; but on attempting this, with the idea that, as usually supposed, they were straight Nautili, it appeared that the narrow aperture, the small outer chamber, the thin outer wall, often apparently only membranous, and the large siphuncle, would scarcely admit of this; and I finished by representing it as something like a modern squid; perhaps wrongly, but it was evidently somewhere between them and the Nautili. Secondly, these synthetic types often belonged to the upper part of a lower group, or to the lower part of an upper group. Hence in one point of view they may be regarded as of high grade, in another as of low grade, and they are often large in size or in vegetative development.[AA] From this law have arisen many controversies about the grade and classification of the Palæozoic animals and plants. [AA] It seems, indeed, as if the new synthetic forms intermediate between great groups were often large in size, while the new special types came in as small species. There are some remarkable cases of this in the plant world; though here we have such examples as the pines and tree-ferns continuing almost unchanged from an early Palæozoic period until now. Thirdly, extinctions of species occur in every great oscillation of the continental areas, but some species reappear after such oscillations, and the same genus often recurs under new specific forms. Families and orders, such as those of the Trilobites and Orthoceratites, appear to have a grand and gradual culmination and decadence extending over several successive periods, or even over the whole stretch of the Palæozoic time. Toward the close of the Palæozoic, while all the species disappear, some whole families and orders are altogether dropped, and, being chiefly synthetic groups, are replaced by more specialised types, some of which, however, make small beginnings alongside of the more general types which are passing away. Our diagram (page 183) illustrates these points.